PSLE Science Reality Lab Vol No.077 | “The Cells Look Bigger” — Were the Two Images Taken at the Same Scale?

PSLE-SCI-REALITY-0077

Wait, What? An object can look twice as large in a picture without becoming any larger at all.

Two microscope images appear beside each other. In the “before” picture, a cell stretches about 2 cm across your screen. In the “after” picture, a similar cell stretches almost 4 cm across. The caption says: “The treatment made the cells twice as large.”

That conclusion might be correct. But the screen size alone is not enough. The second image may have been captured at a higher magnification, cropped more tightly or resized more when the figure was assembled.

A scientific image is not the specimen itself. It is a representation with a scale. To compare physical size, you need to know how image distance maps back to real distance.

Reality Lab Vol No.077 teaches one real-world evidence-transfer job: when two scientific images are used to claim that an object became larger or smaller, check whether the images share a comparable scale before treating visual size as measured size.

Quick Answer

Find the scale bar. What real distance does it represent?
Compare the scale bars, not just the objects. Equal-looking bars may represent different distances; different-looking bars may represent the same distance after resizing.
Check field of view and magnification. A tighter field often makes objects look larger on the page.
Use a real measurement where possible. Compare object length against the scale, or use measurements supplied by the study.
Check selection and conditions. Make sure the images represent comparable specimens and regions.
Limit the conclusion. “Looks bigger in this figure” is not the same as “was physically larger.”

Reality Lab habit: In a scientific image, visible size belongs to the display; physical size belongs to the specimen. The scale connects them.

The Owned Learner Job — and the Boundary

This guide does not become the general owner of magnification, measurement, microscopy or not-to-scale diagrams. Those underlying skills already exist in the eduKateSengkang Science estate.

Vol No.055 asks whether the displayed image was selected fairly from the full experiment. Vol No.077 asks a separate question: even if both images are genuine and representative, can their displayed sizes be compared directly?

Original Reality Lab Case: The Root-Hair Images

Consider two original composite images of root cells. No real paper or commercial source is being copied.

Figure	Cell width on the displayed page	Scale bar on page	What scale bar represents
Image A	24 mm	20 mm	50 µm
Image B	36 mm	30 mm	50 µm

At first glance, the cell in Image B looks 1.5 times wider on the page. But the scale bar is also 1.5 times longer on the page. Relative to the scale, the two cell widths are the same.

Image B was simply displayed larger.

Observed, Claimed and Inferred

Layer	What can be said
Observed	The object occupies more millimetres on the displayed page in Image B.
Claimed	The cell became physically larger.
Inferred	Page size directly represents specimen size without needing scale information.

The observation is about the figure. The claim is about the specimen. The scale is the bridge between them.

Why a Scale Bar Is More Useful Than “40×” Printed Under an Image

A microscope may use a 40× objective, but the final image can later be resized on a screen, in a slide or on a webpage. If the whole image is enlarged, a printed statement such as “40×” does not tell you how many millimetres on your current screen equal one micrometre in the specimen.

A correctly embedded scale bar changes size with the image. That is why image-integrity guidance from the U.S. Office of Research Integrity describes a scale bar of known size as a better way to communicate image scale when the final display size may change.

For a Primary learner, the practical rule is simple: if you want to compare real object sizes in two images, compare each object with its own scale information.

Representation Check: Same Picture, Different Display Size

Take one photograph and place it twice in a document. Make the second copy twice as wide. Nothing in the photographed object has changed, but every feature now looks twice as large on the page.

This trivial example exposes the core problem. A figure is a display. Its visual dimensions can change after the scientific observation is complete.

Therefore “it takes up more space on my screen” is evidence about the display, not automatically about the specimen.

Field-of-View Check: Zooming In Changes What Fits

A lower-magnification image may show twenty cells across a wide field. A higher-magnification image may show only four cells. Those four cells can look much larger because the view covers a smaller real area.

This is useful: scientists zoom in to reveal detail. But a side-by-side “before / after” figure should not invite a physical-size comparison unless the scale makes that comparison valid.

Scale-Bar Check: Equal-Looking Bars Can Mean Different Things

Suppose Image C and Image D each show a 2 cm scale bar on the page. Image C’s bar is labelled 100 µm. Image D’s is labelled 25 µm.

The bars look equally long, but they represent different real distances. An object spanning one bar in Image C is about 100 µm across; an object spanning one bar in Image D is about 25 µm across.

Never compare the drawn bar length without reading its label.

Measurement Check: Convert the Picture Into a Ratio

You do not need advanced microscopy mathematics. Use a ratio.

If a cell appears 30 mm wide on the page and a 20 mm scale bar represents 50 µm, then the cell spans 30/20 = 1.5 scale bars. Its approximate real width is 1.5 × 50 µm = 75 µm.

If another cell appears 45 mm wide but its 30 mm scale bar also represents 50 µm, it is again 1.5 scale bars wide: about 75 µm.

One looks larger on the page. The scaled measurements are the same.

Comparison Check: Are We Looking at Comparable Things?

Even after the scale problem is solved, another question remains. Are the images showing comparable specimens, regions and orientations?

Same type of cell or structure?
Same stage of growth?
Comparable position in the specimen?
Same orientation or cross-section direction?
Same measurement definition?
Representative images rather than one unusually large example?

Reality Lab reasoning works in layers. Fixing scale makes size comparison possible; it does not automatically solve sampling or selection.

Worked Case 1: The “Giant Pollen Grain”

Photo A shows a pollen grain 18 mm across with a 10 mm bar labelled 20 µm. Photo B shows a pollen grain 27 mm across with a 15 mm bar also labelled 20 µm.

Both grains span 1.8 scale bars. The larger-looking grain in Photo B is not evidence of a physically larger grain.

Worked Case 2: Same Display Size, Different Real Size

Two circular structures both appear 25 mm across. In Image E, the 25 mm scale bar represents 100 µm. In Image F, the 25 mm scale bar represents 20 µm.

The structures look the same size on the page, but one spans a real distance of about 100 µm and the other about 20 µm. Equal display size can hide a fivefold real-size difference.

Worked Case 3: No Scale Bar

A before-and-after infographic shows two microscope images with no scale information. The after image is larger on the page. Can you conclude the cells grew?

Not from visual size alone. You would need the acquisition scale, a scale bar, calibrated measurements or another reliable statement of physical size. The images may still show other useful features, but this particular size claim is under-supported.

Worked Case 4: Measured Size Supports the Picture

Two images use the same stated scale and include scale bars. A blinded measurement of 30 cells per group reports a larger mean cell width after the treatment. Now the visual impression is backed by quantitative evidence from multiple specimens.

The claim is stronger—not because the second picture looks dramatic, but because scale, measurement and sampling align with the conclusion.

Image Processing Is a Separate Question

Brightness, contrast, colour mapping and sharpening can affect how a scientific image looks. Reality Lab Vol No.066 already owns the job of evaluating enhancement and possible processing artifacts. Vol No.077 does not repeat that work.

Here the focus is geometric scale: how much real distance does the displayed image represent?

What Evidence Would Strengthen “The Cells Became Larger”?

Scale bars with clear units on both images.
Acquisition information showing comparable imaging conditions where relevant.
Measurements made relative to the calibrated scale rather than page size.
Several cells or fields measured rather than one striking example.
Comparable specimen type, location and orientation.
A stated measurement rule such as width, area or diameter.
Raw or minimally processed data retained where image analysis is involved.

What Would Weaken It?

No scale information.
Different magnifications presented as if page dimensions were directly comparable.
One image cropped much more tightly than the other.
A scale bar removed during figure assembly.
Only one selected cell per condition.
Screen measurements compared without converting through the scale.
Different orientations that make the same structure appear wider or narrower.

Model and Measurement Limits

A scale bar does not make an image perfect. Image resolution, focus, specimen preparation, perspective and the way boundaries are chosen can affect measurement. Some structures are three-dimensional while the image is a two-dimensional section or projection.

For Primary 5/6, you do not need to master these specialist problems. The important habit is to understand what the scale fixes and what it does not. It converts image distance into real distance; it does not guarantee the specimen selection, image quality or biological explanation.

How Far Can the Conclusion Travel?

If ten measured cells from one sample are wider after a treatment, you may have evidence about those measured cells and that sample. A broader claim about every cell, every organism or every condition needs broader evidence.

Scale solves one representation problem. Scientific scope still matters.

PSLE-Style Transfer Case

A student is shown two photographs of the same type of cell. In Photograph X, a cell is 3 cm wide and the 1 cm scale bar represents 20 µm. In Photograph Y, a cell is 4 cm wide and the 2 cm scale bar represents 40 µm. The student says the cell in Y is larger because it takes up more centimetres on the page.

Evaluation: The conclusion is not supported by page size alone. In X, the cell spans three scale bars, giving about 60 µm. In Y, the cell spans two scale bars, also giving about 80 µm—so Y may indeed be larger, but the correct reason comes from the scale conversion, not from the raw centimetres on the page.

The key transfer is that even when the first impression happens to point in the right direction, the reasoning still has to use valid evidence.

Tempting Reasoning That Fails

“Bigger on the page means bigger in reality.” The display may have been resized.
“Both images say 40×, so I can measure them with a ruler.” Final display size can change after capture.
“The scale bars look the same length.” Read what real distance each bar represents.
“The after image shows fewer cells, so each cell must be larger.” A narrower field of view can show fewer cells without any growth.
“A scale bar proves the whole figure is trustworthy.” It supports spatial measurement; other evidence-quality questions remain.
“No scale bar means the image is useless.” It may still show shape, presence, position or other features; it is weaker for physical-size claims.

Explained Practice

Practice A: Two insect-wing photographs are printed at different sizes. Each includes a 5 mm scale bar. What should you compare first? The wing length relative to its own scale bar, not the printed centimetres alone.

Practice B: Two cells occupy equal widths on a screen, but one image’s scale bar represents 100 µm and the other 25 µm. Are the cells necessarily equal in real size? No.

Practice C: An infographic removes scale bars to make images cleaner. What claim becomes harder to check? Any claim about absolute or comparative physical size.

Delayed Independent Return: The BAR Test

Later, when you see a pair of scientific images, remember three letters:

B — Bar: Is there a scale bar, and what does it represent?
A — Acquisition/display: Were magnification, crop or display size different?
R — Real measurement: What physical size does the object have after using the scale?

Then add the usual Reality Lab question: how far can that measurement be generalised?

Parent and Tutor Teaching Guide

You can teach this without a microscope. Photograph the same coin twice. Crop the second image tightly and enlarge it. Place a simple paper ruler beside the coin before photographing so that both images contain scale information.

Ask the learner which coin is larger. After the obvious answer, remove the physical coin from view and ask them to prove it from the images. The learner should discover that the display can change while the real object remains unchanged.

Then transfer to leaves, printed diagrams or microscope images. The educational goal is not calculation speed. It is refusing to use a visual dimension as scientific data until the representation tells you what that dimension means.

Authoritative Sources

ORI’s guidance emphasises that pixels in scientific images have a scale and that scale bars are especially useful when images may be resized for display. That specialist guidance connects cleanly to the Primary Science habit of interpreting representations and evaluating information before drawing conclusions.

The Quiet Return

Scientific images are powerful because they let us see things that may be tiny, distant, faint or otherwise difficult to inspect directly. Their power comes with a responsibility: remember that the picture is a representation.

When one cell, crystal, leaf structure or material grain looks bigger than another, do not begin with the ruler on your screen. Begin with the scale bar.

Then measure the evidence in the same units as the claim.

Small Group Tutorials